Page 191 - Encyclopedia of Philosophy of Language

may be thrown on the nature of mental and semantic concepts. The tactic has a broadly verificationist inspi- ration: their idea is that one can find out what meaning and the mind are, by seeing how one detects them in others.
1. W. V. O. Quine
Use of the scenario for this purpose first came to prominence with the publication in 1960 of Quine's Word and Object. The discussion in ch. 2 of 'radical translation' is a landmark in the philosophy of language, and all debate about the nature of meaning since is necessarily informed by, even if it rejects, Quine's approach. Quine subscribed to behaviorism about meaning, holding that, insofar as it is fixed at all, 'meaning is a property of behavior.' Accordingly, he set out to consider how much of ordinary semantic notions can be constructed from a basis of purely physical facts about the 'natives' disposition to verbal behavior. His true concern was thus not so much epistemic as metaphysical: not to see how we might in practice seek to determine the meanings of native sentences, but rather to explore how, as David Lewis has put it, 'the facts (about behavior) determine the facts (about meaning).' Quine used the scenario as a heuristic device to explore this latter question; although his verificationist leanings mean that for him the metaphysical question of if and how the facts are fixed becomes one with the question of how they might, at least in principle, be verified. Quine con- sidered how a translator might arrive at a correct 'translation manual' from the natives' language into her own. He concluded, notoriously, that the data of natives' dispositions to verbal behavior does not suffice to narrow the choice down to just one: all the constraints available on the translator's task leave many manuals equally acceptable. The most dis- turbing element in this thesis of the 'indeterminacy of translation' was Quine's claim of the 'inscrutability of reference': there is no basis to discriminate between alternative 'analytical hypotheses' about sentences which assign different references to terms and predi- cates, where these yield logically equivalent trans- lations for whole sentences. So, in Quine's example, it is indeterminate whether native talk is about rabbits, rabbit parts, or rabbit time-slices. This thesis about reference has been convincingly argued against by Evans, but it is now generally recognized that some
considerable indeterminacy in translation exists.
2. DonaldDavidson
Quine was concerned solely with how the meaning of native sentences might be discovered by a translator. But it is now generally recognized that this task can be accomplished only simultaneously with another: the 'interpretation' of the speakers of the language to be translated—that is, the ascription to them of beliefs, desires, and other mental states. The impossi-
bility demonstrated by Quine of constructing sentence meanings from facts about speakers' dispositions to verbal behavior is part of the more general falsity of behaviorism. Behaviorism is false because there is no simple, one-by-one relation of the mental states of persons to their observable behavior: what a person does in response to a given stimulus depends not just on what she believes, but also on what she wants, and there is no principled limit on the further mental states which may crucially affect her response. Similarly, what the sentences of a subject's language mean has no implications for her behavior except as mediated by her mental states. Thus Davidson, continuing the investigation of mental and semantic concepts by means of the radical interpretation scenario, noticed how meaning and belief 'conspire' together to deter- mine which sentences a subject holds true (and hence which she will assent to). In Davidson's work the primary focus switches to the mental: his concern is with how a 'radical interpreter' might ascribe mental states to the natives. He holds that the essential nature of the mind can be illuminated by this method. He uses it to argue, for example, that beliefs are by nature mainly true. One must, he claims, use a principle of 'charity' in interpreting others—that is, ascribe to them mainly true beliefs; and he makes a characteristic interpretationist move from this claim about the inevi- table method of interpretation, to a conclusion about the nature of belief itself. The product of a successful interpretation exercise will be both an ascription of beliefs, etc., to the natives, and a theory of meaning for their language. Davidson argues that, while explicit reduction of sentence meaning to non-semantic notions is impossible, by giving an account of how such an interpretation of a community can be achieved, one gives all that is needed by way of philo- sophical explanation of the nature of meaning. He holds that a theory of truth can serve as a theory of meaning, and has suggested that telling the radical interpretation story can also serve as all that is needed by way of philosophical explanation of what truth is. It has, however, been questioned whether the same story can illuminate both truth and meaning.
3. State of the Art
That meanings cannot be constructed from speakers' dispositions, as Quine showed, is now generally recog- nized. But while Quine drew the moral that ordinary semantic notions are not scientifically respectable, most nowadays would conclude instead that his stan- dard for respectability was too severe. But the appeal to radical interpretation in the philosophical eluci- dation of the mental and semantic remains much in evidence. Doubts about Davidson's work focus on two main issues. It is uncontroversial that an interpret- ation must 'make sense' of the individual(s) to be interpreted, and that this requires seeing a certain pattern in the interrelations amongst their mental
Radical Interpretation
169